tajo-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jihoon Son <jihoon...@apache.org>
Subject Brief Evaluations of Tajo on Swift
Date Fri, 27 Feb 2015 05:01:32 GMT
Hi guys,
I have conducted some brief performance evaluations of Tajo on Swift, and
would like to share the results.

I conducted two kinds of experiments; The first experiment was to compare
the performance of Tajo with on another distributed storage, i.e., HDFS.
And the second experiment was the scalability test of Swift.

Interestingly, the scan performance on Swift is slower more than two times
than that on HDFS. In addition, the task scheduling time on Swift is much
greater than that on HDFS, which means the query initialization cost is
very high. You can find the detailed results at the following link.

http://www.slideshare.net/jihoonson/apache-on-tajo-bringing-sql-to-the-openstack-world

Based on those evaluation results, I would like to add some new features to
improve the performance of Tajo on Swift. For example, progressive task
scheduling can mitigate the query initialization cost on Swift. In
addition, we need to support location-aware computing for segmented Swift
objects. Sooner or later, I will create issues on Jira.

Best Regards,
Jihoon

Mime
View raw message